People living with HIV in communities with universal access to HIV testing and treatment were 10% more likely to be employed than those receiving care according to national guidelines, a randomised study has shown. They were also 10% less likely to seek healthcare, 13% less likely to spend money on healthcare and their children were 7% more likely to complete primary school.
Published in the January issue of The Lancet Global Health, this is the first study to use an experimental design to look at whether access to universal testing and treatment results in better social and economic outcomes at the individual, household and community levels. Thus, we can now say with more certainty that universal testing and treatment for HIV is a cause of better social and economic outcomes in rural communities.
The Sustainable East Africa Research in Community Health (SEARCH) trial in Kenya and Uganda was a large study comparing outcomes among communities who had universal access to HIV testing and treatment and those who received the standard of care according to national guidelines. Although it did not achieve its primary aim of reducing HIV incidence, there were promising results in terms of multiple health outcomes, including population-level viral suppression, reductions in vertical transmission, fewer deaths, reductions in tuberculosis, and improvements in hypertension control.
Beyond these important health benefits, Dr Aleksandra Jakubowski from the University of California, Berkeley and colleagues used the randomised cluster experimental design from SEARCH to explore the potential economic and social benefits of universal test and treat.
While there already was evidence to suggest that increases in employment and income, and household benefits more generally – such as improved schooling and nutrition – are associated with adults living with HIV starting antiretroviral therapy (ART), many of these studies were conducted when most people started ART at a later stage of disease, resulting in more dramatic improvements. Additionally, the design of these studies meant that although associations were revealed, these associations could not necessarily be attributed directly to the initiation of ART.
Between 2013 and 2017, 32 rural communities of approximately 10,000 people each from distinct geographical areas in southwestern and eastern Uganda, and western Kenya, were assigned to either intervention or control groups. The communities were matched in 16 pairs in terms of geographical location and characteristics, with one from each pair being randomly assigned to either condition.
At baseline, all communities received multi-disease health campaigns that included services such as screening for HIV, hypertension and cervical cancer; referrals for male medical circumcision; and de-worming treatment for children. Any participants testing positive for HIV during this time were linked to care. These campaigns were repeated annually in the intervention communities only.
In intervention communities during the trial, participants diagnosed with HIV could start HIV treatment immediately, using a patient-centred care model which also included access to services for diabetes and hypertension. This model incorporated three-monthly visits, flexible hours, reduced wait times and staff training to ensure welcoming services.
Participants in control communities received necessary interventions based on the national standard of care guidelines. For HIV, these guidelines were based on CD4 counts at diagnosis. This changed during the study period: initially a CD4 count below 350 cells was required to start treatment, but this later changed to a CD4 count below 500 cells.
Researchers also collected socioeconomic data annually from a subsample of approximately 200 households in all communities (half with at least one adult living with HIV at baseline, half with none). Surveys included information on employment, income, education, ownership of goods and livestock, and use of healthcare. Data for this analysis came from the surveys at baseline and after three years of the SEARCH trial.
At baseline, intervention and control communities were similar: there were a total of 4,053 adults aged 18 to 65 years in the intervention communities (4,136 in the control communities), and 3,217 children aged 6 to 17 years (3,538 in the control communities). Over half were female (57%); 43% had completed some primary school and 74% were classified as employed (as measured by having done any work in the week before the survey).
Of the adults in both communities, 34% were living with HIV, of whom 48% had a CD4 count above 500. Approximately 20% had sought out healthcare services for illness or injury and had spent money on healthcare in the previous month. The average age of children was 11, approximately half were female and 93% attended school. However, only around 8% had completed primary school.
There were no meaningful or significant differences in employment and use of healthcare between HIV-negative participants living in intervention and control communities after three years.
This was not the case for people living with HIV: adults in the intervention communities reported were nearly 10% more likely to be employed (an increase of 9.7%; 95% confidence interval 2.1 to 18.3) and were 10% less likely to use healthcare services after the three-year trial (a decrease of 10.3%; 95% CI -22 to 0.1). As could be expected, those in the intervention communities received multi-disease interventions and were 12.7% less likely to spend money on health expenses (95% CI -22.4 to 0.6).
Interestingly, the effects of the intervention were more pronounced in those who had higher CD4 counts at baseline. Those with CD4 counts above 500 were 21.6% more likely to be employed (95% CI 5.3 to 40.4) and spent around 13 more hours a week on work. Additionally, those with higher CD4 counts in intervention communities were 13.5% (95% CI -31.5 to 4.2) less likely to lose time from usual activities because of illness than those with high CD4 counts in control communities.
School-age children in households with an adult living with HIV were 7.3% (95% CI 1.2 to 14.8) more likely to finish primary school three years after the intervention than those in control communities. This difference rose to 10.8% (95% CI 0.6 to 22.7) if the adult had a CD4 count above 500. Results were much more pronounced when only children over the age of 12 were included in the analysis, as they were 15.8% (95% CI 3.8 to 26.3) more likely to finish primary school in intervention communities.
“The effects of the intervention were about twice as large among adults with high CD4 count (>500 cells) at baseline, which suggests that maintaining good health through early ART initiation is a crucial mechanism by which economic benefits accumulate,” the authors conclude. “Children who lived with HIV-positive adults had higher educational attainment in intervention communities than in control communities. These findings are consistent with the hypothesis that early ART provision to HIV-positive people not only protects their health but also their economic wellbeing and the human capital outcomes of children in their households.”
“The results from this study indicate that universal HIV testing and ART provision combined with streamlined care delivery generate substantial socioeconomic benefits for individuals and households in rural Africa,” they add.
Jakubowski A et al. Effect of universal HIV testing and treatment on socioeconomic wellbeing in rural Kenya and Uganda: a cluster-randomised controlled trial. The Lancet Global Health 10: e96-e104, 2022 (open access).
Full image credit: HIV testing. Villagers line up to get tested for HIV in a makeshift laboratory in Uganda. Image by Victoria Holdsworth/Commonwealth Secretariat. Available at www.flickr.com/photos/comsec/3989910492 under Creative Commons licence CC BY-NC 2.0.